Effects of Thermal Fatigue on Nitriding

Hot Working Die Steel (H13)

Koji Yatsushiro

and Masahiko Hihara

Yamanashi Industrial Technology

Center,

Kofit, Yamanashi 400 Japan

and
Katsuhiro Tagawa and Makoto Kuramoto
Department

of Mechanical Engineering,

Polytechnic University, Sagamihara,

Kanagawa 229-I 1 Japan

Abstract
X-ray stress constants

of nitriding layers on a hot working

determined for different types of nitriding processes.

die steel (AISI

H13) were

Thermal fatigue behaviour was investigated

by

the thermal fatigue test and the heating test.

The results obtained are as follows: (I) The X-ray stress constant of nitriding specimen A (E Fez_sN and y -Fe4N, treated

at 843K for 6h) was KA= -648 MPa/deg.

for the E - Fez.sN 103

diffraction peak using Cr-K, radiation, and for specimen I3 (E - FesN and y-FedN, treated at 833K
for 30h) was KB = -920 MPa/deg.
(2) Nitriding

specimens

showed

for the E -Fe3N 103 diffraction

a decrease

of heating resistance

peak using Cr-K, radiation.

due to decomposition

of the

nitriding layer during the heating test. The oxidation of the nitriding layer began at 723K (450C) as
examined by measuring weight, X-ray diffraction profiles and by microphotographic

observations.

1. Introduction
Die casting dies and forging dies are considerably

damaged by thermal stresses which are

repetitions of heating and cooling). One of the ways to increase the life of a die is to use a nitriding
process, such as gas-nitriding,

carbo-nitriding,

tufI%-iding and plasma-nitriding.

They are widely used

in industrial products.
After the nitriding process is applied to a hot working
thermal fatigue characteristics
However,

die steel, it shows an increase of

because of the effect of compressive

during the thermal cycles, the heating temperature

(6OOC), the nitrided layer decomposes

stress and a hardened

is about from 773K (500C) to 873K

and nitrogen gas is released from the surface. Then the

Copyright (C) JCPDS-International Centre for Diffraction Data 1997

thermal fatigue characteristics

surface2).

decrease as the number of thermal cycles increases.

Thermal

fatigue behaviour

of nitriding

hot working

die steel had been investigated

by

measuring residual stress using a Fe 211 diffraction with Cr-K, radiation including the peaks of Fe-N
compounds,

and by observation

of diffraction profiles and crack growth3. It is considered that it will

be a useful method to realize the thermal fatigue phenomena

residual stress using single peaks of Fe-N compounds.

However,

compound

layers on the die steel have not been determined.

concerning

X-ray stress measurement

and on pure iron. Furthermore,

of nitriding die steel by measuring

X-ray stress constants

The few data published

of nitriding layers are carbon&riding

the phenomena

of Fe-N
so far 4,5)

layers on carbon steel

of nitriding layers applied to die castings and hot

working forging dies during operation have not been clarified.

The purposes
constants

of this study are as follows: (1) Determination

for two types of nitriding specimen.

(2) To examine the thermal fatigue phenomena

nitriding layers during the thermal fatigue test (heating-cooling

stress and X-ray diffraction
layers

in the

heating

microphotographic

profiles. (3) Investigate

test

by

measuring

of the X-ray stress and elastic

the

cycle) by measurements

the decomposition
weight,

X-ray

and oxidation

diffraction

in

of X-ray
of nitriding

profiles

and

by

observations.

2.

Experimental

2.1

Determination

procedure
of X-ray stress constants

Specimens used are shown in Table 1. The specimens were machined by means of grinding.
Before the nitriding process, specimen B was heat treated at 923K(650C)

for 5 hours in argon gas

for the purpose of stress release.

X-ray difIi-action profiles of the specimens are shown in Fig. 1. Diffraction plane analysis was
employed for identification

of the nitrides. For specimen A, the nitrides were E -Fez_jN*) and y -

Fe4N. Also, a Fe was identified. For the specimen B, the nitrides were E -Fez_sN, E -Fe3N and y Fe4N. Fez03 was formed during the stress release heat treatment.
From the results of identification for specimen A, E -Fez_sN 103 diffraction by Cr-K, radiation
was used. The peak position

was about

134deg. Also, for specimen

B, &-Fe3N 103 and 200

diffraction with the same radiation was used. The apparatus for X-ray stress measurement
RIGAKU

MSF-2M

which has a parallel beam

system

and w -goniometer.

was a

It was used for

There
are two cards for Centre
&-phase
(No.Data
30925
and No. 11236) in the JCPDS file and both of them are described
Copyright *(C)
JCPDS-International
fornitride
Diffraction
1997
as Fe3N. However, the card No. 30925, had been described as Fe2_3N.In this study, in order to differentiate one from
the other, the card No. 30925 is described as Fe2_3N,and No. 11236 as Fe3N.

measurements of the E -Fe2_3N 103 diffraction

goniometer

and the E -FesN 103 diffraction.

was used for the E -FesN 200 diffraction. In the measurement

the specimens were bent by a four-point-bending

Also, an R -

of X-ray stress constants,

device with a strain gauge. X-ray stress and elastic

constants were obtained from change of slope of 2 0 -sir? v diagrams. A strain gauge was attached
to the tension
specimens

side which is the opposite

one to the X-ray diffraction.

were from 0 to 600x 10m6with an interval

Strains applied to the

of 100x 1Om6,and 800~10-~ was added.

Especially, in case of the 103 diffraction for the specimen A and the 200 diffraction for the specimen
B, a strain value of 1OOOxlO-6was added.
five times.

The stress measurements

for each specimen were repeated

Applied stresses in this experiment were calculated using the Youngs modulus for H13

(208.8 GPa). The values of sin* w in the diagram were from 0 to 0.6 with an interval of 0.1.
Table1 Specimens

used in this study

B

Specimen
Material

Width-length-thickness,mm

Heat treatment and hardness

Nitriding condition

Hot working die steel (AISI H13,JIS

SKD61)

10-100 -3
Quenching and tempering

,45 HRC

833K(560C),30h

843K(570C),6h

600
P
g 400
E

40

60

-40

80

Diffraction angle,

28

(deg.1

(a) Nitriding specimen A

60

80

100

Diffraction angle, 2 e

120

140

160

(deg.)

(b) Nitriding specimen B

Fig. 1 X-ray diffraction pr-ofiles of the specimens

2.2 Thermal fatigue test
Specimens

in the form of cylinders (58mm diameter and 20mm thickness)

were prepared

from hot working die steel (AISI H13, JIS SKD61) which was heat treated and hardness of 48HRC
was obtained.

The nitriding process was performed under a gas atmosphere

Copyright (C) JCPDS-International Centre for Diffraction Data 1997

hours.

at 793K (520C) for 20

In the thermal fatigue test, the specimen was heated up to 843K (570C) for 160sec. by
contact to a heating block, and was cooled to 373K ( 100C) in a water bath. Heating temperatures
are representative

of aluminum die casting operation.

1000 cycles (repetition of heating and cooling).

the surface were measured

The thermal fatigue test was carried out up to

During the thermal fatigue test, residual stresses on

step by step using the X-ray stress constant

value obtained from the

results mentioned in the paragraph below.

An X-ray diffraction profile of the specimen for the thermal fatigue test is shown in Fig.2.
Diffraction

from the E -Fezs3N 103 for which the stress constant

was -648MPaldeg.

with Cr-K,

Also, the peak at about 156deg. which was diffracted from both the E - Fez_jN

radiation was used.

200 and the a Fe 211 was used with the X-ray stress constant (-297MPa/deg.)

Diffraction angle,

28

of the hot die steel.

(deg.)

Fig. 2 X-ray diffraction profile of specimen for the thermal fatigue test

2.3 Phenomena

of oxidation during heating

Specimens were heat treated to 45HRC and prepared by wire EDM (15mm width, 25mm
length and 0.5mm thickness).

After removal of the eroded layer by emery paper, nitriding was

performed at 843K (560C) for 30 hours, same as for specimen B.

Specimens

were heated from 673K to 973K, for each 50K, they were kept 6 hours then

cooled in the furnace.

Specimens were evaluated by observation

weight ratio, Vickers Hardness and X-ray diffraction profiles.

heated at temperature

of sectional area, measurement

The weight ratio A WT for specimens

T was obtained from the equation (1)

A W,

= w;WQ
0

x 100

(%)

(1)

Copyright where
(C) JCPDS-International
for Diffraction
1997 before and after heating.
W0 and WT are Centre
the weight
of the Data
specimen

Copvrinht 0 JCPDS-International

Centre for Diffraction Data 1997

of

3. Results and discussion

3.1 Determination

of X-ray stress constants

Fig,3 shows the CJx-M and CJx-2 8 V=~diagrams obtained from each specimen, where M is the
slope of the 2 0 -sin2 w diagram, 2 8 ,+,=ois the intercept, and o x is the applied stress.
the diagram of specimen B for the E - FesN 103 diffraction were somewhat scattered.
other results:

The plots in
However, the

E Fe2_3N 103 diffraction for specimen A and E -FesN 200 diffraction for specimen B

indicated good linearity.

The peak from the E -Fe2_3N 103 diffraction plane is about one deg. higher than the peak used
and might influence determination
the E -FejN 200 diffraction
determination

of the peak position.

is more accurate

Stress measurement

than that from

for specimen B, using

103 diffraction.

In case of the

of the X-ray stress constant of the E -FejN where the 103 diffraction peak overlaps the

E -Fez_sN 103 diffraction as in specimen B, it is necessary to separate the peaks.

?
P

APPII~
,._

stress,o r (MW

(a) Specimen A for the E -Fe2_3N 103

134.20 2
3

-P -0.10
3
r
g -0.20
0
B
3
5 -0.30
ii

..:.....
r ........

162.8

pzh,,,i
.............!
_.._
/2Ip

L....

.................
,

..,+

&.

..........

j.. ............

;/; i
A*::.
: .....
...... i.. ........... . ...... ;_.
j

....

i..

...........

... .

I- ...

50

150

Applied stress, CJy

162.7

e
3

50

100

150

200

250182.6

200
(MPa)

(b) Specimen B for the E -FesN 103 peak

(c)

Fig.3 B x -M and CJx -28

Copyright (C) JCPDS-International Centre for Diffraction Data 1997

Centre for Diffraction Data 1997

8
2
S

d
t
Applied stress, a y

CowriaRtO JCPDS-International

t3

.: ..

0
100

162.8
i ..

5 -0.40
0

(MPa)

Specimen B for the E -FesN 200 peak

v=~

diagrams

The values of X-ray stress and elastic constants obtained from these figures and the residual
stresses are shown in Table 2.
compressive

The confidence

stresses were accumulated

limit shown in the table is 50%.

in specimen A.

It is found that

Also in specimen B, tensile stress were

obtained.
Table 2

X-ray stress constants of nitriding specimen A for Fez_sN 103 and

specimen B for FesN 103 and the 200 diffraction peaks
tress constant

=--.

s, L-E

828,

3 (.Tx

.-

dM

2(1w)
. a

fjo

s,

=--

--

18)

r
*-

3 (J x

a% &

1%)

3.2 Thermal Fatigue Test

Fig.4 and Fig.5 show the change of residual stress and peak intensity during the thermal
fatigue test. The confidence limit shown in Fig.4 is 50%.
Before the test, a compressive
diffraction peak.
convex shape.

stress of -3OOMPa was observed

The residual stress profile measured

from the E -Fez_sN 103

from the E -Fez_sN 103 difiaction

had a

The intensity of this diffraction peak began to decrease at 50 cycles and disappeared

at 500 cycles.
On the other hand, before 20 cycles, the residual stresses obtained from the peak at 156 deg.
namely the E -Fez_sN 200 diffraction with the a Fe 211 showed the same results as the E -Fe2_3N 103
diffraction.

At 50 cycles and 100 cycles, it was observed that the residual stress tended to decrease

and was scattered widely. After 200 cycles, the residual stress fell in the range of about -200 to
-4OOMPa. The value of the peak intensity increased 2 times at 1000 cycles, however, that of the E Fe2.3N 103 diffraction

decreased

between

20 and 100 cycles, that of the overlapped

peak didnt

change up to 100 cycles.

Fig.6 shows the half-value breath of both of the peaks during the thermal fatigue test.
half-value breath of the E -Fe2_3N 103 diffraction

decreased

slightly, that of the overlapped

The
peak

around 156 deg. fell off to nearly 3.3 deg. which is the value for quenched and tempered die steel.
From the results
peak intensity
Copyright (C) JCPDS-International
Centreofforresidual
Diffractionstress,
Data 1997

and half-value breath measurements,

at

around 10 to 20 cycles some changes were begun to be observed in values from both the E -Fe2_3N

103 and the overlapped

peaks.

In the range from 100 to 500 cycles, major inflection points exist in

each case. In this experiment, because the heating temperature

to the nitriding temperature,

these phenomena

pm depth of X-ray penetration.

correspond

in the thermal fatigue test was close

to decomposition

of nitrides in the 5 to 10

The existence of unstable crystals due to decomposition

might cause

the change of the values and the wide scatter. For the overlapped peak, more crystallization of the
a- Fe 211 diffraction might stabilize the values in the range over 500 cycles.
processes applied to dies, during thermal cycling (heating-cooling),

In case of nitriding

it is hard to check with the eye

what the condition of the nitriding layer is. It became clear that X-ray stress measurement

using the

peak from the E -Fez_sN 103 diffraction and the a Fe 211 with the E -Fez_sN 200 diffraction would be
one of the useful methods of quality control.
to be separated.

However, when the overlapped

peak is used, it needs

Also, the peak separation and the influence on residual stress of the decomposition

of the nitride layer should be considered.

400 :

+Fe,,N2OO+aFe211

100
10
Number of cycles, Nf (cycle)

1000

I. ..I
-+-FeJ4

.
2OO+aFe211
103

10

. ...I

* . !
1
i...."".'._'.~""......~
,

100

1000

Number of cycles, N, (cycle)

Fig.6

Changes in half-value breath during the thermal fatigue test

Copyright (C) JCPDS-International Centre for Diffraction Data 1997

Cnnvrinht Q .ICPiI.~lntt=mdionA

1000

Fig. 5 Peak intensity of the specimen

during the thermal fatigue test

_f_._.__._
-a--h,$J

100

Number of cycles, N, (cycle)

Fig. 4 Residual stress during the thermal

fatigue test
?

10

Cmtre for Wfraction Data 1997

3.3 High temperature oxidation behaviour

Microphotographic observations of cross sectional area near the layers produced by the
nitriding process and after heating are shown in Fig.7(a), (b) and (c). Also, the X-ray diffraction
profiles are shown in Fig.8.
The nitride layer shown as a white color was observed at the surface of the specimen, as
produced. The thickness of the layer was about 5 to 10 pm. Below the layer, deposition which is
referenced6 as iron carbide (it is called snake like carbide) was observed at grain boundaries. This
deposition existed along grain boundaries up to about 50 pm underneath the nitrided layer, at more
than 50 pm, it was parallel to the surface. From the results of X-ray diffraction, E -Fe3N, Y-FeAN, a
few a Fe and Fez03 peaks were observed.
After heating at 723K(450C), brittle and porous oxide layers were formed at the surface of
the nitride, whose thicknesses were about 5 pm. The appearance of Fe203 peaks was observed in the
X-ray diffraction profiles.
After heating at 673K(400C), the oxide layers were not observed in microphotographs
Fe203 peaks were evidenced by X-ray diffraction measurement. Also, changes in surface color of the
specimen were observed. These results coincide with the result of reference7) on the oxidization of
nitrided (y-Fe4N) prepared from pure iron.
AfIer heating at 823K(550C), analysis of the X-ray diffraction peaks, E -FesN, y-FeJN,
Fe203 and new a Fe peaks were identified. Around this temperature, the nitride layer decomposed
and nitrogen gas was released from the surface. On the other hand, due to low activity of oxygen,
the decomposed nitride layer was observed as a Fe. In this experiment, the penetration depth of CrK, radiation in a Fe is about 5 to 10 pm for the 110 diffraction peak (2 8 =69deg.), it does not
diffract through the layer. These results coincide with the results of the previous section.

~~

~~~~~~

(a) Before heating

--

(b) Heating at 723K(450C)

Fig.7

(c) Heating at 823K(550C)

Microphotographic observation of cross sectional areas

Copyright 0 JCPDS-International Cenbe for Diffraction Data 1997

TqplEr

* 0 Fe A Fez03 x FmN . FmN

Temp. I

700
600
500
400
80

60

40

100

120

140

160

1.

1.

100

150

1.

250

300

200

Distance from surface fi m

Diffraction angle ,2 6 (deg.)

Fig. 9 Distribution curves of hardness

Fig. 8 X-ray diffraction profiles

Hardness distribution

50

curves for the various nitriding specimens are shown in Fig.9.

In this

figure, the result for the specimen before heating is shown with 0 mark.
Hardness

did not change

at heating

temperature

decreased about 20HV at 823K(550C) due to formation

on the surface.

less than 773K(500C),

however,

of oxides after decomposition

it

of nitrides

Also, the decrease below the nitride layer was due to the effect of tempering

and

dif-&sion of solute nitrogen.

Hardness

decreased

in all of the region at 873K(600C),

furthermore,

it decreased

in the

range from the surface to a depth of about 50 pm. At 973K(700C), the hardness fell about 500 to
550 HV in all of the range, because of the influences of vigorous decomposition
X-ray microanalyzer
about 16%.

and difI%sion. The

@MA) results are shown in Fig. 10. Nitrogen content of the nitride layer was

Because of this the nitride is similar to the one formed by the tufRriding process)

Before heating, the high nitrogen content area was observed to be about 40 urn. Below this
area, it decreased with the distance from the surface, and became equal to the matrix at about 250
urn. After heating at 873K (6OOC), a high oxygen content

area with formation

of oxide was

observed in the region near the surface. On the other hand, the distribution curve of nitrogen content
had a convex shape. This curve corresponded

to that of nitrogen.

Copyright (C) JCPDS-International Centre for Diffraction Data 1997

Copyright@ JCPDS-International

Centre for Diffraction Data 1997

Distance from surface

100

200

300
/L m

Distance from surface

p m

(b) Heating at 873K(600C)

(a) Before heating

Fig. 10 XMA line analysis of nitrided specimens

Also, after heating at 873K(600C), oxidation progressed

nitrides, hardness was maintained by solute nitrogen in the difision
The relationship between heating temperature

at the surface by decomposition

of

layer.

and weight ratio calculated from equation (1)

is shown in Fig. 11.

400

500
600
700
Temperature, T (c)

Fig. 11 Relationship between weight ratio and heating temperature

An increase of weight was observed
oxides on the die steel specimen.
inactivity at high temperature

at more than 923K(650C),

This specimen

because

(H13) is the hot working

by forming chromium oxides.

Due to the protection

of the surface

steel which shows

effect on forming

of oxide on the surface, an increase of the weight ratio for the die steel specimen was not observed
Copyright (C) JCPDS-International Centre for Diffraction Data 1997

for T < 873K(600C).

On the other hand, an increase of the weight ratio for the nitride specimen

for T > 723K(450C).

was observed

In the nitriding

process,

preference to other elements due to its high affinity for nitrogen.

content

of chromium,

the heating resistance

nitrides such as Me(Fe,

Cr), decompose

of the specimen

enlargement

binds to nitrogen

Because of the reduction

is decreased.

in

in the

At high temperature,

and are released from the surface, nitrogen

redif&sed in the diffusion layer, and at the same time imperfections,

and grain boundaries.

chromium

atoms are

i.e., pores are formed at surface

It is considered that as oxidation progresses, the weight ratio increases due to

Therefore,

of surface area by the formation of imperfections.

a nitriding process, as in

this experiment, applied to a die casting die, is able to be one of the useful methods of increasing die
life such as the reduction of the maximum temperature

at the die surface.

4. Conclusion
Determination

of X-ray stress constants

different types of nitriding

processes

of the nitriding layer on hot working

were carried out and thermal

die steel for

fatigue behaviour

has been

investigated by thermal fatigue tests and the heating tests. The results obtained were as follows:

0)

X-ray stress constant

of nitriding

for 6h) was KA= -648MPa/deg.

specimen

A (E -Fe2_3N and y-FehN, treated

with the E -Fez_sN 103 diffraction

at 843K

peak of Cr-K,

and

specimen B (E -FesN and y-FedN, treated at 833K for 30h) was Kn = -920 MPa/deg. with
the E -FesN 103 diffraction peak of Cr-K,.

(2)

In the thermal fatigue test, however, peak intensity from the E -Fez_sN diffraction decreased
with the number of the thermal cycles and disappeared

at 500 cycles, while that from the E -

Fez_sN with the a Fe around 156 deg. increased with the number of the cycles.

(3)

Nitriding specimens

showed a decrease of heating resistance

nitriding layer during the heating test.

(450C)

as

examined

microphotographic

by

due to decomposition

of the

The oxidation of the nitriding layer began at 723K

measuring

weight,

X-ray

diffraction

profiles

and

by

observations.

Acknowledgments

We are indebted to K. Kiyoyama of Nichiei Corporation

M. Sano of Yamanashi Industrial Technology

Center and M. Egawa of Polytechnic

ray stress measurement.

Copyright (C) JCPDS-International Centre for Diffraction Data 1997
Copyright8 JCPDS-International

for preparing specimens.

Centre for Diffraction Data 1997

We thank

University for X-

References

1)

M. Hihara

and Y. Mukoyama,

Engineering,

2)

International

Journal

of The Japan

Society

for Precision

Vol.25 No.3, 187 (1991).

A. Schindler, A. Kulmberg and J.H. Stuhl, 9th Society of Die Casting Engineers Congress, GT77-065 (1977).

3)

M. Hihara and Y. Mukoyama,

Journal of The Japan Society for Precision Engineering,

Vo1.57,

No.6, 1005 (1991)(In Japanese).

4) K. Tanaka, G.Q. Ping, T. Mikuriya, Y. Akiniwa, Proc. of the 3 1st Workshop on X-ray Studies
of Mechanical

Behaviour

of Materials, The Society of Material Science, Japan, 55 (1994)(In

Japanese).

5) M. A.J. Somers and E. J. Mittemeijer, Metallurgical Transactions. A,2 1A, 189, (1990).

6)

S. Ohira, Y. Osada, Proc. of the 87th conference

of the Japan Institute of Light Metals, 337

(1994)(In Japanese).

7)

M.A.J. Somers and E.J. Mittemeijer, Metallurgical Transactions.

8)

M. Hihara, Y. Mukoyama and I. Ogata, Journal of The Japan Society for Precision Eng. Vo1.56,
No.8, 1508 (1990)(In Japanese).

Copyright (C) JCPDS-International Centre for Diffraction Data 1997

Copyright 0 JCPDS-International

Centre for Diffraction Data 1997

A,21A,901, (1990).